We propose to investigate relations between fertility, mortality, and development in Brazil from 1960 to 2000. We will analyze the fertility transition in both time and space, using microdata from five censuses carried out during the period. Our investigation both follows and extends recent analyses of European historical experience, and will attempt to answer fundamental questions about the nature and determinants of the demographic transition in a large developing country. The length, size and detail of the Brazilian census microdata series is, we believe, unsurpassed among developing countries. Moreover, it enables us to study a period covering the bulk of the transitions in fertility and mortality for this large and diverse population: from 1960 to 2000 Brazil's Total Fertility Rate fell from 6.0 to 2.3, while 5Qo fell from about 170 to 45 per 1000. The specific aims of the project are to construct and map over 500 spatial units called micro-regions that can be identified in and compared across all five censuses; estimate fertility and mortality for each micro-region in each census; construct indicators of various theoretically relevant aspects of development and living conditions in the rural and urban areas of the micro-regions in each census; estimate fixed effects models to answer questions about the degree to which changes in these indicators are associated with changes in fertility with refinements to account for endogeneity and local feedback; and estimate the timing and speed of the fertility and mortality transitions in each of the micro-regions using Bayesian methods to fit logistic curves for each micro-region's time series. The overall goals are to assess whether and how the threshold for the initiation of the transition shifted during the 1960-2000 period, as well as the degree to which development is associated with the fertility transition in these spatial units.